Assessment of Common Rights Metadata Encoding Schemes and Gap Analysis

Introduction

The descriptive elements that are needed to support a copyright analysis are similar to elements that are generally used in bibliographic description. This is important for two reasons: first, it means that some of the elements may be already available in existing bibliographic record formats; and second, we may be able to re-use data input for bibliographic description in the copyright description. The actual data elements that we need to provide for in our copyright metadata are:

• Creator, personal
• Creator, corporate
• Creator (unspecified)
• Creator death date
• Copyright holder
• Date of publication
• Date of copyright
• Date of creation
• Date of renewal
• Publisher
• Country of publication
• Published/unpublished
• Type of resource

In addition, we will need some textual fields for:

• Contact information: This can be the contact information for the rights holder or for the institution that can negotiate rights or inform the user about the rights status of the item.
• Other terms and conditions: Until we have a comprehensive set of data elements for license and contract agreements, there does need to be a field that can carry a brief statement of terms and conditions that can be displayed to users. This will not be an actionable field, and should refer to the contact information for those who wish to make specific uses of the item.

Comparison to Bibliographic Description

Below are some areas of difference between the information needed to support the description of copyright status and the data elements and coding practices used for bibliographic description.

Unknown vs. No Attempt to Code

Minimal level cataloging has a set of required fields, such as author and title. The items being digitized in some projects will not necessarily get that minimal level of cataloging, especially in situations where a large number of items are treated in a batch. It is important to know when particular data elements are missing from the original piece (i.e. an item with no named creator), or when the contributing library or archive has not provided the information even though it may be present. Therefore there is a difference between "no author" and "no attempt to code" (in MARC21 parlance). There is no way to indicate this for the main descriptive fields in DC, MODS, or even MARC21.

Rationale: The user needs to know whether there is potentially information available on the original item that may be relevant to its copyright status, or if all relevant information has been included in the digital metadata.

Published vs. Non-Published

Other than the presence of a publisher name in the metadata, there is no way to indicate that the item being described is unpublished as defined in copyright law. Therefore there is often no difference between an item that is unpublished and a published item where the name of the publisher is not supplied.

Creator

The creator is a key element in copyright law, even if the creator is no longer the copyright holder. Information about the creator is necessary for the expiration of the rights over time. Copyright law has different rules relating to personal creators vs. corporate entities. This distinction is not available in Dublin Core. It can be expressed in MODS and MARC21.

Also, the key date in relation to a person as creator is the death date, since that is the date that determines the length of the copyright. The death date can be expressed in MODS and MARC21, although it doesn't have its own explicit mark-up. For example:

1903-1969 ; d. 1969

However, in library cataloging, dates are only included with author’s names when those are needed to distinguish between two authors with the same name, and only the date of birth is needed for that purpose. Occasionally, the only known date is the death date, and therefore that is the date used, with "d." before it to indicate that it is the date of death. In practice, records with author’s date of birth are rarely updated to include the date of death, so this key piece of information is missing from most cataloging.

Dublin Core does not have a data element for the creator’s dates.

Resource Dates

There are four possible relevant dates relating to a resource:

• Date of creation
• Date of publication
• Copyright date
• Date of renewal

Because the dates of creation and publication can be unknown, there needs to be a way to give approximate dates, or even ranges of dates, such as a century or decade.

For an unpublished item, the date of creation is the key date. In the world governed by the Berne Principles, the item is copyrighted at the time of creation, so the date of creation is essential to understanding the copyright status of the item. MARC21 does not have a creation date other than the "date of manufacture" subfield in the 260 ($g). MODS translates this "date of manufacture" to its "dateCreated" field. MODS also has "dateIssued" which receives most of the dates encoded in the MARC21 coded date element from the 008 field. These include known and approximate dates, as well as copyright and publication date.

Dublin Core has a general "date" data element that may be used for a variety of dates. It also has "dateCreated," "issued" (i.e. published) and "dateCopyrighted." There is no express way to code approximate dates.

One final date used by copyright law is the date of copyright renewal This date applies to copyrighted works during the time period that renewals were possible. (Copyright renewal in the U.S. ended in 1976.) None of the metadata schemes has a data element for this date.

Country of Origin or Publication

Copyright law can very depending on the country in whose jurisdiction the item was created or published. MARC21 and MODS can provide the "place" (generally place of publication) in a fixed field (MARC21 008-15/17; MODS <place>). Dublin Core does not have a data element for place of publication.

Type of Resource

US copyright law specifically references the following types of resources:

• Literary works
• Musical works
• Dramatic works
• Choreographic works
• Pictorial, graphic and sculptural works
• Motion pictures and other audiovisual works
• Sound recordings
• Architectural works

As the table below shows, each metadata scheme has a different set of resource types based on their own point of view. MODS uses the resources from MARC21, which have their basis in library cataloging and in the separate bibliographic formats developed as the MARC formats. Dublin Core resource types have some overlap with the MARC21 formats, although some distinctions are missing. For example, MARC21 recognizes "cartographic materials," and Dublin Core does not. Presumably, maps would be treated as "images" in Dublin Core. This may reflect Dublin Core’s emphasis on digital materials, where items like maps are electronically indistinguishable from other images.

Copyright holder

Standard library catalog does not specify the copyright holder, nor does it record a copyright statement found on the piece. This lack is carried through to MARC21 and MODS. Dublin Core does have a specific field for "rightsHolder" although in the extension to its element set. This element is important for copyright determination because although the duration of copyright is based on the author’s life span, the actual rights may be held by the publisher of a work, or by another party.

Table: Comparison of Dublin Core, MODS, and MARC with Copyright Description Elements

Data element	Dublin Core	MODS	MARC	US Copyright Law
Creator, personal	No	Yes	Yes	Yes
Creator, corporate	No	Yes	Yes	Yes
Creator (unspecified)	Yes	No	No	No
Creator death date	No	No	Possible, but rare	Yes
Copyright holder *	Qualifier: rightsHolder	No	No	Yes
Date of publication	Qualifier: issued	Not specifically	Not specifically	Yes
Date of copyright	Qualifier: copyrightDate	Not specifically	260 $c "c. 1999"	Same as date of creation?
Date of creation	Qualifier: created	Not specifically	Not specifically	Yes
Date of renewal	No	No	No	Yes
Publisher	Yes	Yes	Yes	Only if copyright holder? Otherwise, may be assumed to be on deposited item.
Country of publication	No	Yes	Yes (008)	Yes. Country of first publication of the work
Published/unpublished	No	No	No	Yes
Type of resource	Collection Text Sound Image Still image Moving image Physical object Software Event Interactive resource Service	Manuscript Collection Text Cartographic Notated music Sound recording – musical Sound recording – non-musical Still image Moving image 3-d object Software Mixed media	Language material Notated music Manuscript notated music Cartographic material Manuscript cartographic material Projected medium Nonmusical sound recording Musical sound recording Two-dimensional nonprojectable graphic Computer file Kit Mixed material Three-dimensional artifact or naturally occurring object Manuscript language material	Literary works Musical works Dramatic works Choreographic works Pictorial, graphic and sculptural works Motion pictures and other audiovisual works Sound recordings Architectural works

* The use of METSRights would allow this data element to be added to the descriptive data covered by any of the schemes listed here.